The cast pigs, or blocks of black or crude copper, are piled upon the hearth, each successive layer crossing at right angles the layer beneath it, in order that the flame may have access to play upon the surface of the hearth, and to heat it to a proper pitch for making the metal flow.
The weight of the charge should be proportional to the capacity of the furnace, and such that the level of the metallic bath may be about an inch above the nozzle of the bellows; for, were it higher, it would obstruct its operation, and were it too low, the stream of air would strike but imperfectly the surface of the metal, and would fail to effect, or would retard at least, the refining process, by leaving the oxidation and volatilization of the foreign metals incomplete.
As the scoriæ form upon the surface, they are drawn off with an iron rabble fixed to the end of a wooden rod.
Soon after the copper is melted, charcoal is to be kindled in three iron basins lined with loam, placed alongside the furnace, to prepare them for receiving their charge of copper, which is to be converted in them, into rosettes.
The bellows are not long in action before the evaporation of the mineral substances is so copious, as to give the bath a boiling appearance; some drops rise up to the roof of the reverberatory, others escape by the door, and fall in a shower of minute spherical globules. This phenomenon proves that the process is going on well; and, when it ceases, the operation is nearly completed. A small proof of copper, of the form of a watch-case, and therefore called montre, is taken out from time to time, upon the round end of a polished iron rod, previously heated. This rod is dipped two or three inches into the bath, then withdrawn and immersed in cold water. The copper cap is detached from the iron rod, by a few blows of a hammer; and a judgment is formed from its thickness, colour, and polish, as to the degree of purity which the copper has acquired. But these watches need not be drawn till the small rain, above spoken of, has ceased to fall. At the end of about 11 hours of firing, the numerous small holes observable in the first watch samples begin to disappear; the outer surface passes from a bright red to a darker hue, the inner one becomes of a more uniform colour, and always less and less marked with yellowish spots. It will have acquired the greatest pitch of purity that the process can bestow, when the watches become of a dark crimson colour.
Care must be taken to stop this refining process at the proper time; for, by prolonging it unduly, a small quantity of cupreous oxide would be formed, which, finding no oxygen to reduce it, would render the whole body of copper hard, brittle, and incapable of lamination.
The basins must now be emptied of their burning charcoal, the opening of the tuyère must be closed, and the melted copper allowed to flow into them through the tap-hole, which is then closed with loam. Whenever the surface is covered with a solid crust, it is bedewed with water; and as soon as the crust is about 11⁄2 inch thick it is raised upon hooks above the basin, to drain off any drops, and then carried away from the furnace. If these cakes, or rosettes, be suddenly cooled by plunging them immediately in water, they will assume a fine red colour, from the formation of a film of oxide.
Each refining operation produces, in about 12 hours, 17⁄10 tons of copper, with the consumption of about 4⁄5 of a ton of dry wood.
Care should be taken that the copper cake or rosette be all solidified before plunging it into water, otherwise a very dangerous explosion might ensue, in consequence of the sudden extrication of oxygen from the liquid metal, in the act of condensation. On the other hand, the cake should not be allowed to cool too long in the air, lest it get peroxidized upon the surface, and lose those fine red, purple, and yellow shades, due to a film of the protoxide, which many dealers admire.
When a little oxide of antimony and oxide of copper are combined with copper, they occasion the appearance of micaceous scales in the fractured faces. Such metal is hard, brittle, yellowish within, and can be neither laminated nor wire-drawn. These defects are not owing to arsenic, as was formerly imagined; but, most probably, to antimony in the lead, which is sometimes used in refining copper. They are more easily prevented than remedied.